This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Introduction: Only recently has it been explored that some dyslexic readers may classify as 'resilient readers'because of their average or above average text comprehension skills. Specific Aims: To investigate the neural basis of resilient readers, Methods: A total of 62 healthy native English speakers between 6-16 years of age (mean 13.4, standard deviation (SD) 2.6) participated in our study. Among them, 31 children were skilled readers and 31 were dyslexic readers with and without poor reading comprehension skills (15 and 16 children, respectively). We performed voxel-based morphometry (VBM) and examined differences in regional gray and white matter volume (GMV, WMV) in dyslexic adolescents with and without poor reading comprehension. We also performed phonological and semantic processing tasks utilizing functional magnetic resonance imaging (fMRI) to identify whether there were corresponding differences in brain activation. Results: Whole brain analysis of GMV regressing out total GMV showed significantly greater GMV in the left parieto-temporal region in the Dys[NormComp] compared to the Dys[PoorComp] Group (p<0.01 corrected). This cluster included two distinct peaks in the left posterior superior temporal gyrus (pSTG) and inferior parietal lobule (IPL). There were no significant differences in WMV. Post-hoc region-of-interest analysis showed significantly greater GMV in controls compared to Dys[NormComp] and Dys[PoorComp] combined (t(59)=2.09, p=0.04), but not compared to Dys[NormComp] only (t(59)=0.72, p=0.48). Corresponding to the cognitive characteristics, left parieto-temporal activation during rhyme judgment showed no significant difference between Dys[NormComp] and Dys[PoorComp], but Dys[NormComp] compared to Dys[PoorComp] showed significantly greater activation during semantic categorization in this region.